Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Fri, 16 Jun 2023 16:39:24 +0100
From:   Sudeep Holla <sudeep.holla@arm.com>
To:     Tushar Nimkar <quic_tnimkar@quicinc.com>
Cc:     "Rafael J . Wysocki" <rafael@kernel.org>,
        Sudeep Holla <sudeep.holla@arm.com>,
        Daniel Lezcano <daniel.lezcano@linaro.org>,
        Ulf Hansson <ulf.hansson@linaro.org>, linux-pm@vger.kernel.org,
        linux-arm-msm@vger.kernel.org, linux-kernel@vger.kernel.org,
        quic_lsrao@quicinc.com, quic_mkshah@quicinc.com,
        devicetree@vger.kernel.org
Subject: Re: [PATCH 1/2] dt-bindings: arm: idle-states: Add
 idle-state-disabled property
Message-ID: <20230616153924.2wtvgr7lvjcul6to@bogus>
References: <20230608085544.16211-1-quic_tnimkar@quicinc.com>
 <20230608085544.16211-2-quic_tnimkar@quicinc.com>
 <20230615085629.b2aaumhq7yqhs5lf@bogus>
 <5820345a-4207-3b12-87eb-098bac4ef4e8@quicinc.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <5820345a-4207-3b12-87eb-098bac4ef4e8@quicinc.com>
Precedence: bulk

On Fri, Jun 16, 2023 at 11:26:18AM +0530, Tushar Nimkar wrote:
> 
> Thanks for review Sundeep,
> 
> On 6/15/2023 2:26 PM, Sudeep Holla wrote:
> > On Thu, Jun 08, 2023 at 02:25:42PM +0530, Tushar Nimkar wrote:
> > > +      idle-state-disabled:
> > > +        description: |
> > > +          If present the idle state stays disabled. It can be enabled back from
> > > +          shell using below command.
> > > +          echo N > /sys/devices/system/cpu/cpuX/cpuidle/stateX/disable
> > > +        type: boolean
> > > +
> > 
> > This is clearly a policy and not a hardware or firmware feature to expose
> > in the device tree. So NACK, why can't you load it modules if you don't want
> > idle states in the boot.
> > 
> Attempt of making cpuidle governors to modular was rejected in past [2]
>

OK try command line approach to disable all states(you can't get partial
on/off in that case). I don't think the build config is of any use as we
end up enabling it which will affect other platforms.

> [2] https://lore.kernel.org/lkml/1637830481-21709-1-git-send-email-quic_mkshah@quicinc.com/#t
> 
> > It is same as choosing any default governor or performance states, will you
> > add those next ? It is simply policy not a feature/property to be exposed
> > in the device tree.
> > 

The above still holds, so still NACK. It is a policy and not a
hardware/firmware property or feature.

-- 
Regards,
Sudeep